DE102014116904B4 - Method for optically scanning and measuring a scene and automatically generating a video - Google Patents

Abstract

A method for optically scanning and measuring a scene by means of a laser scanner (10), which generates at least one scan of the scene with a specific center (Ci) and with several measurement points (X), wherein the generated scans with their measurement points (X) in a common Coordinate system (XYZ) of the scene are registered and the totality of all measurement points (X) of the generated scans form a three-dimensional point cloud (3DP), which is visualized, wherein the automatic generation of a video as visualization a) in a first step in the common coordinate system (XYZ ) the scene is defined a trajectory (T) with multiple trajectory points (TP) on it by placing excellent points in a sequence and defining exact or approximated trajectory points (TP) connected together according to the sequence, b ) in a second step, the trajectory points (TP) clusters (G ', G °) are assigned, each cluster (G', G °) ei represents a sequence of trajectory points (TP), and wherein for the allocation of the trajectory points TP to the clusters (G ', G °) the curvature of the trajectory (T) at the trajectory points (TP) is determined, a trajectory point (TP) belongs to a "straight" cluster (G ') if the curvature is below a certain threshold in absolute value, and at least two consecutive trajectory points (TP) form an "extended" cluster (G ° ), if the curvatures amount in each case above the determined threshold and have the same sign, c) to the clusters (G ', G °) assigned trajectory points (TP) in each case a section of the local environment (αβγ) from the Three-dimensional point cloud (3DP) is determined, resulting from a juxtaposition of these sections, the video.

Description

The invention relates to a method for optically scanning and measuring a scene by means of a laser scanner ( 10 ), which generates at least one scan of the scene with a specific center (C _i ) and with several measurement points (X), wherein the generated scans with their measurement points (X) are registered in a common coordinate system (XYZ) of the scene and the totality of all Measuring points (X) of the generated scans a three-dimensional point cloud ( 3DP ), which is visualized.

In the DE 10 2009 015 922 A1 a method of this kind is described in which a scene with multiple scans is detected. For this purpose, the laser scanner is moved to a new location after a scan in order to generate another scan. The generated scans are registered with their measuring points in a common coordinate system, the totality of the measuring points forming a three-dimensional point cloud.

The visualization of this three-dimensional point cloud takes place by means of a two-dimensional projection from a selected viewpoint in a selected viewing direction. It is customary to select a subarea of the three-dimensional point cloud and transfer it to the graphics card, which then generates the two-dimensional projections itself. Due to the size of the three-dimensional point cloud, it is not possible to get an overview with random points of view and random viewing directions. In addition, even with the selection of a subarea of the three-dimensional point cloud, huge amounts of data still have to be moved.

The post-published DE 10 2013 110 580 A1 and WO 2015/048048 A1 describe a method in which a trajectory is laid through the three-dimensional point cloud, along which an interactive video is presented as visualization along the user, with a more selective view into the local environment and, if appropriate, choice at branching points.

The invention is based on the object to improve a method of the type mentioned. This object is achieved by a method with the features of claim 1. Advantageous embodiments are the subject of the dependent claims.

The inventive method achieves a significant data reduction by a data preprocessing of the three-dimensional point cloud, without any significant restrictions for the user. The data reduction not only allows a faster data transfer and visualization of the remaining data, but the generated video also makes it possible to get a (meaningful) overview of the three-dimensional point cloud. Accordingly, the method has two intentions. On the one hand, the scan project is to be supported by finding and recognizing deficits of the three-dimensional point cloud, such as data poor or difficult to visualize or faulty areas, so that additional scans can be performed from other locations and errors can be eliminated. On the other hand, the overall result of the scan project should be visualized by showing as much as possible, in particular important and well-recorded areas.

For the video, a fixed trajectory is defined, in particular by connecting trajectory points obtained from the various centers occupied by the laser scanner during the various scans and from other excellent points, such as entrances, exits and orbits around obstacles, to avoid collisions of the trajectory and the data. Variations in altitude are reduced or eliminated. The formation of clusters and the definition of at least one focus for each cluster allows the definition of preferred viewing directions. The associated section of the local environment is preferably adapted to the human visual field. The sequence (sequence) of the excerpts forms the video. The speed of movement, which results from the number of intermediate points assigned to the cutouts on the trajectory, can be selected dynamically.

In the following the invention with reference to an embodiment shown in the drawing is explained in more detail. Show it

1 a perspective view of the laser scanner,

2 a schematic representation of the laser scanner in operation,

3 a partial schematic representation of components of the laser scanner,

4 a planar view of a scan,

5 a panorama view of a scan,

6 a schematic representation of the scene with three-dimensional point cloud and trajectory,

7 a schematic representation of the three angles of rotation at a trajectory point,

8th a side view of the laser scanner showing the surroundings of a trajectory point,

9 a flowchart for generating a video,

10 the determination of a trajectory in a first step,

11 the allocation of trajectory points to clusters in a second step,

12 the definition of a focus in each cluster in a third step, and

13 the definition of a visible section of the local environment of each trajectory point in a fourth step.

The present invention relates to a 3D coordinate measuring device which directs a light beam onto an object O which is either a (cooperative) target, for example a retro-reflector, or a non-cooperative target, for example a diffusely scattering surface of the object O, can be. A distance meter in the device measures a distance to the object O, and rotary encoders measure the rotation angles of two axes in the device. The measured distance and the two angles allow a processor in the device to determine the 3D coordinates of the object O. In the present case, such a device becomes the case of a laser scanner 10 treated, but the extension to a laser tracker or a total station is obvious to the expert.

Laser scanners are typically used to detect closed or open spaces, such as building interiors, industrial facilities and tunnels. Laser scanners are used for many purposes including Building Information Modeling (BIM), industrial analysis, accident reconstruction applications, archaeological surveys and forensic investigations. A laser scanner can be used to optically detect and measure objects in the laser scanner environment by capturing data points that represent objects within the environment. Such data points are obtained by directing a light beam at the objects and collecting the reflected or scattered light to determine the distance, two angles (i.e., one azimuth angle and one zenith angle), and optionally a gray tone value. This raw scan data is collected, stored and sent to one or more computers to generate a three-dimensional image representing the captured area or object. To generate the image, at least three values are collected for each data point. These three values may include the distance and two angles, or may be converted values such as x, y, z coordinates.

The drawing shows a laser scanner 10 for optically scanning and measuring the surroundings of the laser scanner 10 , The laser scanner 10 has a measuring head 12 and a foot 14 on. The measuring head 12 is so on the foot 14 mounted that the measuring head 12 around a first axis 12a relative to the foot 14 is rotatable, driven by a first rotary drive. The rotation about the first axis 12a can be around the middle of the foot 14 respectively. The measuring head 12 has a mirror 16 on which is a second axis 16a can rotate, powered by a second rotary actuator. Relative to a normal, upright position of the laser scanner 10 , may be the first axis 12a be referred to as a vertical axis or azimuth axis, and the second axis 16a can be referred to as a horizontal axis or zenith axis. The laser scanner 10 may have a gimbal point or center C ₁₀ , which is the intersection of the first axis 12a and the second axis 16a is.

The measuring head 12 further comprises a transmitter for electromagnetic radiation, for example a light emitter 17 on, which has a transmission beam 18 sending out. In the preferred embodiment, the transmitted light beam is 18 a coherent light such as a laser beam. The laser beam may have a wavelength in the range of about 300 to 1600 nm, for example 790 nm, 905 nm, 1550 nm, or less than 400 nm. In principle, however, other electromagnetic waves with a larger or smaller wavelength can be used. The transmitted light beam 18 may be amplitude modulated or intensity modulated, for example with a sinusoidal or rectangular waveform. Alternatively, the transmitted light beam 18 be otherwise modulated, for example, by a chirp signal, or coherent reception methods can be used. The transmitted light beam 18 is from the light emitter 17 on the mirror 16 given, deflected there and in the surroundings of the laser scanner 10 sent.

A reflected light beam, subsequently as a received light beam 20 is reflected by an object O in the environment. The reflected or scattered light is from the mirror 16 captured and onto a light receiver 25 redirected with a receiving optics. The directions of the transmitted light beam 18 and the receiving light beam 20 arise from the angular positions of the measuring head 12 and the mirror 16 around the axis 12a respectively. 16a , These angular positions in turn depend on their respective rotary actuators. The angle of rotation about the first axis 12a is detected by a first rotary encoder. The angle of rotation around the second axis 16a is detected by a second rotary encoder.

A control and evaluation device 22 stands with the light transmitter 17 and the light receiver 21 in the measuring head 12 in data connection, with parts of the control and evaluation device 22 also outside the measuring head 12 can be arranged, for example as a foot 14 connected computer. The control and evaluation device 22 is designed for a large number of measuring points X a corresponding number of distances d between the laser scanner 10 and the measurement points X on the object O to determine. The distance to a certain measuring point X is at least partially determined by the running speed of the light in the air through which the electromagnetic radiation propagates from the device to the measuring point X. In the preferred embodiment, the phase shift is in the modulated light beam 18 . 20 which is sent to and received from the measurement point X is determined and evaluated to obtain a measured distance d.

The speed of light in air depends on air properties such as air temperature, air pressure, relative humidity and carbon dioxide concentration. These air properties affect the refractive index of the air. The speed of light in air is the speed of light in vacuum divided by the refractive index. A laser scanner of the type described herein is based on the light transit time in the air (the transit time the light takes to travel from the device to the object and back to the device). A method of distance measurement based on the time of flight of light (or the duration of another type of electromagnetic radiation) depends on the speed of light in air and is therefore easily distinguished from triangulation distance measuring methods. In triangulation-based methods, light is emitted from its light source in a particular direction and then captured on a camera pixel in a particular direction. By knowing the distance between the camera and the projector and matching a projected angle with a reception angle, the triangulation method allows the determination of the distance to the object based on a known length and two known angles of a triangle. The triangulation method therefore does not depend directly on the speed of light in air.

The measuring head 12 can be a display device 24 include that in the laser scanner 10 is integrated. To the display device 24 includes a user interface, which may be a graphical touch screen as shown in the drawing. For example, the display device 24 Have a user interface that allows the operator to the laser scanner 10 To give measuring instructions, in particular to set the parameters or the operation of the laser scanner 10 to start, and the display device 24 can also display measurement results.

In one embodiment, detecting the environment around the laser scanner 10 by means of a fast rotation of the mirror 16 around the second axis 16a instead, while the measuring head 12 slowly around the first axis 12a rotates, wherein the assembly moves in a spiral. In an exemplary embodiment, the mirror rotates 16 with a top speed of 5820 revolutions per minute. A scan is defined as the totality of the measurement points X of such a measurement. For such a scan, the center C ₁₀ defines the origin of the local stationary frame of reference. The foot rests in this local stationary frame of reference 14 ,

In addition to the distance d from the center C ₁₀ to a measuring point X, the laser scanner 10 still detect a gray level value with respect to the received optical power. The greyscale value can be achieved, for example, by integrating the bandpass-filtered and amplified signal in the light receiver 21 be determined via a measuring period X associated measurement period. Optionally, color images using a color camera 25 be generated. By means of these color images, it is also possible to assign colors (R, G, B) to the measuring points X as additional values.

Preferably, the mirror 16 planar and formed on a mirror carrier 16b arranged. The mirror carrier 16b has the basic shape of a cylinder around the second axis 16a , where the cylinder is 45 ° to the second axis 16a is cut off. The mirror is corresponding 16 elliptical and at 45 ° to the second axis 16a (and at the same time at 45 ° to the first axis 12a ). Preferably, the light receiver 21 a reception lens 21a on, whose optical axis with the second axis 16a flees. The diameter of the receiving lens 21a corresponds approximately to the small semiaxis of the elliptical mirror 16 , The light receiver 21 preferably also has a mirror arrangement 21b on, in the direction of the receiving light beam 20 seen behind the reception lens 21a is arranged. The mirror arrangement 21b forms the received light beam 20 on the detector 21c starting at the control and evaluation device 22 connected.

The light transmitter 17 preferably has a laser diode 17a with modulator and amplifier connected to the control and evaluation 22 connected. The from the light transmitter 17 generated transmitted light beam 18 is by means of a deflection optics 17b and a beam splitter 17c on the second axis 16a deflected and thus reaches the mirror 16 , Due to the deflection optics 17b can the light transmitter 17 with respect to the second axis 16a laterally to the receiving lens 21a be arranged. Due to the beam splitter 17c can be on the optical axis of the receiving lens 21a that is the second axis 16a , between the beam splitter 17c and the receiving lens 21a , the color camera 25 be arranged, for example in a recess of the receiving lens 21a , The function of the beam splitter 17c is preferably dependent on wavelength (dichroic) and / or polarization and / or beam direction, so that the transmitted light beam 18 redirected the light for the color camera 25 but is allowed through. Because the color camera 25 coaxial to the light receiver 21 can be arranged for the control of the color camera 25 the control of the measuring head 12 around the first axis 12a and the control of the mirror 16 around the second axis 16a be used.

The laser scanner 10 has its own three-dimensional coordinate system for the measuring points X and thus for the scan, which is determined by the distance d and the angles of rotation of the first axis 12a and second axis 16a is defined. There are several possibilities, the three-dimensional scan on a two-dimensional view on the display device 24 map. One possibility is a planar view 80 as they are in 4 is shown. The term "planar view" is understood to mean a view of a three-dimensional, unprocessed scan or image in which the recorded greyscale values (and optionally color values) are arranged in columns and rows as pixels. Accordingly, the two coordinate axes in the drawing plane result from the two angles of rotation, the distance d goes into the depth. Due to the spherical nature of the scan, the pixels near the poles correspond to a smaller solid angle than near the equator, resulting in distortion. Even when mapping a rectangular grid onto a sphere, for example in the Mercator projection, a distortion results. Therefore, actually appear straight lines, such as the railing 81 , in planar view 80 curved. The scan can be processed to match the pixels in the planar view 80 Information about the coordinates will be provided. By means of a cursor 82 a pixel is selected, to which, for example, a display 83 the rows / columns, one ad 84 the angle of rotation and a display 85 Cartesian coordinates are possible. The planar view 80 can be displayed at full resolution or in reduced resolution.

In order to eliminate the distortion, a correction can be made to produce a linear relationship between partial areas of the two-dimensional view on the one hand and solid angles in the three-dimensional scan on the other hand, for which the three-dimensional data is projected onto a sphere or a cylinder. The result is a panorama view 90 as they are in 5 is shown. The panorama view 90 is the two-dimensional view of the three-dimensional scan from a particular perspective. It gives a similar impression as in the human eye or a two-dimensional camera. The panorama view 90 may be limited in terms of angular ranges to a particular section, for example, according to the human field of view. By properly processing the scan can be to the pixels in the panorama view 90 Information about the coordinates will be provided.

The term "panoramic view" is intended to refer to representations in which (for the viewing direction of a viewer) an angular movement about a fixed point in space is possible, especially if only a section is shown, however, a translation in space is not provided. Full SD rendering would allow (for the viewer) angular movement and translation in space from point to point. Both the panorama view and the 3D representation are usually the result of displaying two-dimensional images from different perspectives, which in combination create a 3D effect with the impression of depth for the viewer. When such a combination is shown along a path, it gives the impression of a 3D representation. When such a combination is shown as a rotation, the impression of a panoramic view is created. As a precaution, it should be noted that the term "two-dimensional images" may have different (but related) meanings. In the simplest case, these are images of a camera or an eye. But it can also be projections of a three-dimensional point cloud 3DP act. The two-dimensional images can also be edited.

In order to capture a scene from different directions or an extended space, multiple scans are generated from different locations (corresponding to a set of different centers C _i ) and then registered in a common coordinate system XYZ of the scene. The laser scanner 10 it has to change its location, which then each center C ₁₀ of the laser scanner 10 is moved within the common coordinate system XYZ to a new center C _i from said set. To change the location easily, is the laser scanner 10 preferably mounted on a trolley, as in the DE 10 2010 033 561 B3 is described, so that the laser scanner 10 from a location with a center C _i to the other location with a new center C _i can be driven and at the same time the relative position of the centers C _{i is} detected. Alternatively, the laser scanner 10 mounted on a tripod. When all scans are registered in the XYZ common coordinate system of the scene, the set of all measurement points X of all scans forms a three-dimensional point cloud 3DP ,

In abstract terms, the registration consists of a first step to take a first scan, d. H. a first set of measurement points X with three-dimensional coordinates in a first frame of reference with a first center, from a second step to capture a second scan, d. H. a second set of measurement points X with three-dimensional coordinates in a second frame of reference with a second center different from the first center, and a third step, the two scans, d. H. bring the two sets of measurement points X, in a common frame of reference, namely in the common coordinate system XYZ the scene, for which change the three-dimensional coordinates at least a total of measuring points X.

Around the three-dimensional point cloud 3DP to visualize, according to the invention, a trajectory T is placed through the scene in the coordinate system XYZ. Of the theoretically infinite points of the trajectory T, several trajectory points TP are selected which are (slightly) spaced apart. The local environment αβγ is determined for each of the trajectory points TP, ie a local view of the three-dimensional point cloud from the corresponding trajectory point TP 3DP , The local environment αβγ should be on the display device 24 are shown, preferably in a panoramic view, as described above. In view of this, preferably, the gaps between the nearest measurement points X from the three-dimensional point cloud 3DP filled, ie surfaces defined, as for example in the DE 10 2010 020 925 A1 is described. In the preferred limit case, the trajectory points TP are arranged close to each other according to the resolution used, ie the trajectory T consists only of trajectory points TP.

The respective coordinate system for the local environment αβγ of each trajectory point TP is preferably defined with three (rotational) angles about three axes X ', Y', Z 'of a local coordinate system X'Y'Z'. The roll angle β describes a rotation about an axis X ', which corresponds to the tangent to the trajectory T. The pitch angle γ describes a rotation about the transverse axis Y '. The yaw angle (yaw) α describes a rotation about the (vertical) axis Z '. In the simplest case, the roll angle will be β = 0 °, and the other two angles α, γ will result from the trajectory T, ie their derivatives in the coordinate system XYZ. In this case, as a rule, the trajectory T with respect to the Z axis will run at the same height as the centers C _i , so that the pitch angle γ in the simplest case is also 0 °.

To visualize the three-dimensional point cloud 3DP the trajectory points TP are approached one after the other, and the display device 24 shows for each current trajectory point TP (ie the viewpoint) the local environment αβγ. The of the display device 24 For example, the environment shown may be a complete panorama view by showing everything according to a wrap-around camera. Preferably, however, the display device shows 24 the local environment αβγ only partially, namely preferably limited to the human visual field, ie an angular range of the yaw angle α of about 180 ° and an angular range of the pitch angle γ of about 120 °. There are also wide-angle or fish-eye views selectable.

Basically, it would be possible for the visualization the trajectory T through the three-dimensional point cloud 3DP free to choose and in addition on the display device 24 to change the used viewing direction into the local environment αβγ (and, if applicable, the displayed section). For a number of applications, a simplified visualization of the three-dimensional point cloud is sufficient 3DP , This simplified visualization without interactions (in particular without definition possibilities of the trajectory T, without branches, without changes in the display of the local environment αβγ and without interactive elements) is referred to as video and is generated automatically. The control and evaluation device 22 has a suitable video generator.

The trajectory T of the video is in the coordinate system XYZ at a constant or at best slightly changing altitude and has only in the other two (lateral) coordinates greater changes, which simplifies the automatic generation of the video and "roller coaster rides" avoids. This requirement is met, for example, in buildings where the scans were created on the same floor, even if there are cable bridges, hose bridges, sleepers or other low-height ramps.

In a first step 101 the trajectory T of the video and several trajectory points TP are defined on it. Preferably, this first step begins 101 with different sub-steps, in which first excellent points, the at least approximately on the trajectory T, and whose sequence are defined.

The starting point is the totality of the centers C _i (i = 1,... N) of the scans., Which form excellent points which should as far as possible lie on the trajectory T, at least approximately. In addition, there may be other excellent points, such as doors or windows, so that the trajectory T does not pass through walls, or bypass paths for obstacles or other points of interest at which the trajectory T should pass. These other awarded points may be user defined or automatically calculated from the data.

From this set of centers C _i and the other points of excellence, a sequence for the course of the trajectory T is to be defined. The sequence of the points awarded can be determined automatically by sorting criteria, the sorting criteria preferably being hierarchically ordered. Such sorting criteria are, for example, the coarse spatial distance, so that sorted by buildings, the height so that (inside the building) is sorted by floors, the number of the laser scanner, if working with several laser scanners, and finally, if after the above Sorting criteria is preferably pre-sorted, the timing of the scans (or their inversion) and / or the spatial proximity of the awarded points. In principle, the latter sorting criterion would be sufficient, but with the use of several sorting criteria, an "unnatural" course of the trajectory T is later avoided. In addition, the awarded points are assigned to certain clusters, which may coincide with the clusters G ', G ° mentioned later. The sort criteria include certain thresholds by which the allocation to the clusters is decided. The three coordinates in the coordinate system XYZ can be treated differently. Thus, for a sorting according to buildings, the lateral distance is more important than the height difference, and for the sorting on the floor, the coordinates in the z-direction are to be weighted higher. It is also possible to define for the sequence of points awarded by means of an optimization, in particular to use the approach to the "traveling salesman problem".

One possibility is to define the trajectory T as a spline (polynomial) through the existing points of excellence in the said sequence. The awarded points thus become trajectory points TP. In the lowest order, the trajectory T directly connects the trajectory points TP with their exact coordinates, so that the trajectory is T angular. With the higher orders, the trajectory T becomes smoother.

Another possibility is to define the trajectory T as the approximate connection of the points awarded; H. One, several, or all of the designated points approximately define the associated trajectory points TP on the trajectory T by shifting them slightly. An example would be the use of least squares. Another example would be to take single or multiple coordinates of the distinguished points only approximately for the associated trajectory point TP, and then place the trajectory T as a spline through the trajectory points TP (with the approximated coordinates).

For example, of the three coordinates of the excellent points, in particular of the centers C _i , in the coordinate system XYZ, the actual heights can be weighted less and the resulting heights can be adopted as approximated coordinates for the assigned trajectory points TP. For this purpose, for example, a mean height of all centers C _i can be determined and the respective deviations can be reduced for this purpose, for example with a certain factor, so that only slight fluctuations of the height occur. With regard to the two lateral coordinates in the coordinate system XYZ, the trajectory points TP coincide with the assigned excellent points.

The trajectory points TP preferably have distances as closely as possible along the trajectory T. For the following steps better results are expected. If necessary, therefore, additional trajectory points TP between the already existing trajectory points TP are defined on the trajectory T.

In a second step 102 the trajectory points TP are now assigned to clusters G ', G °, where each cluster G', G ° represents a sequence of trajectory points TP and therefore has a start point and an end point. For this purpose, a curvature of the trajectory T is determined for each trajectory point TP. Depending on the differentiability of the trajectory T, a substitute variable can also be determined, for example, in the case of a polygon, the respective angle between the two edges in front of and behind the trajectory point TP. If the curvature at a trajectory point TP lies below a certain threshold in absolute value, this trajectory point TP with its two neighbors (before and behind it) belongs to a "straight" cluster G '(dash-dotted in FIG 11 ). If the bends are at least two successive trajectory points TP amount in each case above the specified threshold value and have the same sign, these include at least two trajectory points TP to an "extended" cluster G ° (dashed in 11 ). So far in the first step 101 in preparation (larger) clusters were formed, these are taken over and, if necessary, subdivided further.

If the curvature of a previous trajectory point TP on the subsequent trajectory point TP exceeds the threshold value, this trajectory point subsequent TP forms the end point of a "straight" cluster G '. If the curvature drops from a previous trajectory point TP to the following trajectory point TP below the threshold value, this preceding trajectory point TP forms the starting point of a "straight" cluster G '. If the curvatures of a preceding trajectory point TP on the subsequent trajectory point TP are each above a certain threshold value and change the sign, this subsequent trajectory point TP forms the end point of an "extended" cluster G °. The starting point of the trajectory T has the same angle as the second trajectory point TP. Each endpoint of a cluster also forms the starting point of the next cluster. Optionally, these clusters are "straight" clusters G 'with only two trajectory points TP. As a result, each trajectory point TP belongs to at least one cluster G ', G °, and successive clusters G', G ° each overlap exactly one trajectory point TP.

Instead of the assignment by means of the curvature of the trajectory T, other criteria can also be used. Likewise, other types of clusters are conceivable, for example with widely spaced trajectory points TP, or outside or inside a building.

In a third step 103 For each cluster G ', G ° a focus F is defined. For each "extended" cluster G °, the focus F is preferably the center of gravity of the cluster G °, ie an average of all the trajectory points TP contained in the cluster G °. For each "even" cluster G ', the focus F is preferably the end point of the cluster G'. In each case different definitions of the focus F are possible. It is also possible that one or more clusters G ', G ° have more than one focus F. For example, areas of the three-dimensional point cloud are suitable as the focus F. 3DP with particularly low dot density or non-closed surface if supplementary scans are to be done there, with particularly high density of dots when presentations are to be made, or with defects in the three-dimensional point cloud 3DP to detect registration or record errors. The focus F can also be an object O, more precisely a centeral point thereof. A definition by the user is conceivable.

In a fourth step 104 the visible section of the local environment αβγ is defined for the given points on the trajectory T, ie for the first step 101 defined trajectory points TP and for intermediate points, which are defined now.

Like the trajectory points TP, each intermediate point is defined by a position in the three-dimensional coordinate system XYZ. At the same time, each intermediate point on the trajectory T lies between two tranjectory points TP. So many intermediate points are defined that later at least twenty-five per second can be approached. The intermediate points can be distributed uniformly along the trajectory T, or depending on trajectory parameters of the trajectory T (for example, a high density of intermediate points if the trajectory T is curved, and a lower density of intermediate points if the trajectory T is even), or depending on the data (for example, a high density of intermediate points if the three-dimensional point cloud is close to the current trajectory point TP, and a lower density of intermediate points if it is farther away). The intermediate points are preferably selected so that the later traversing of the trajectory T is perceived as "pleasant" as possible. This is achieved, for example, by making changes in the density of the intermediate points or, in general, the speed along the trajectory T continuous, as well as changes in viewing directions.

In order to define said section of the local environment αβγ, the predetermined points are each assigned a preferred viewing direction. Within each cluster G ', G °, at least the yaw angle α is chosen such that as a rule the viewing direction (indicated by arrows in FIG 13 ) is set to the focus F of the cluster G ', G °, regardless of the course of the trajectory T. At the starting point of the cluster G', G °, the viewing direction is preferably towards the focus F of the cluster G ', G °. For example, between the penultimate trajectory point TP of the cluster G ', G ° and the end point of the cluster G', G °, the line of sight may be towards the end point of the cluster G ', G °, while at the end point of the cluster G', G °, which is also the starting point of the next cluster G ', G °, the viewing direction preferably goes to the focus F of this next cluster G', G °. If a new viewing direction is to be set, this is preferably done between the preceding trajectory point TP and the one with the new viewing direction, preferably not faster than the average speed when trajectory T is traversed in the video. If necessary, no further points on the trajectory T are approached until the new viewing direction is set. The viewing directions at the intermediate points are preferably interpolated. The pitch angle γ is preferably selected to match the described viewing direction, ie changes in height during the trajectory T are compensated. Alternatively, the pitch angle γ = 0 ° is constant. The roll angle β is preferably always β = 0 °. Just as the points on the trajectory T are interpolated, the view direction is also interpolated.

In this fourth step 104 becomes at any given point of the trajectory T in the three-dimensional point cloud 3DP viewed in the set viewing direction and to the two-dimensional image, which forms the visible section of the local environment αβγ, as a projection (in the present embodiment, on the image plane of the display device 24 ). The visible section is preferably a panoramic view 90 , which is preferably adapted to the human visual field. By juxtaposing the thus obtained sections of the local environment αβγ results in the video, which preferably has the desired twenty-five (or more) frames per second.

For each of these two-dimensional images of the video preferably remain the coordinates of the given point from which is viewed, and the set viewing direction stored, so the camera parameters for the visible section of the local environment αβγ. With these camera parameters, a seamless change between the video and a three-dimensional representation is possible. The user may preferably interrupt the video at any point of the trajectory T, from which point the local environment αβγ may be considered, for example the three-dimensional point cloud 3DP or a panoramic view of the local αβγ environment, and then optionally resume the video.

In a modified embodiment, two slightly spaced camera positions are assigned to the predetermined points on the trajectory T for a stereoscopic viewing, so that in each case two images are generated. It then creates a stereoscopic video with different images for each eye.

LIST OF REFERENCE NUMBERS

10

laser scanner

12

probe

12a

first axis

14

foot

16

mirror

16a

second axis

16b

mirror support

17

light source

17a

laser diode

17b

deflecting

17c

beamsplitter

18

Transmitted light beam

20

Reception light beam

21

light receiver

21a

receiving lens

21b

mirror arrangement

21c

detector

22

Control and evaluation device

24

display device

25

color camera

80

planar view

81

railing

82

cursor

83

Display of rows + columns

84

Display of the rotation angle

85

Display of the Cartesian coordinates

90

panoramic view

101

first step

102

second step

103

Third step

104

fourth step

3DP

(three-dimensional) point cloud

C ₁₀ , C _i

Center of the laser scanner

d

distance

F

focus

G'

straight cluster

G °

extended cluster

O

object

T

trajectory

TP

Trajectory point

X

measuring point

XYZ

(common) coordinate system

X'Y'Z '

local coordinate system

α

yaw

β

roll angle

γ

pitch angle

αβγ

local environment

Claims (8)

Method for optically scanning and measuring a scene by means of a laser scanner ( 10 ), which generates at least one scan of the scene with a specific center (C _i ) and with several measurement points (X), wherein the generated scans with their measurement points (X) are registered in a common coordinate system (XYZ) of the scene and the totality of all Measuring points (X) of the generated scans a three-dimensional point cloud ( 3DP A) in a first step in the common coordinate system (XYZ) of the scene, a trajectory (T) with multiple trajectory points (TP) is defined on it by excellent points in a sequence and define exactly or approximated trajectory points (TP) which are connected together according to the sequence, b) in a second step the trajectory points (TP) are assigned to clusters (G ', G °), each cluster ( G ', G °) represents a sequence of trajectory points (TP), and wherein for the allocation of the trajectory points TP to the clusters (G', G °) the curvature of the trajectory (T) at the trajectory points ( TP), where a trajectory point (TP) belongs to a "straight" cluster (G ') if the curvature is below a certain threshold in absolute value, and at least two consecutive trajectory points (TP) become a " extended "clusters (G °) belong, if the curvatures amount in each case above the certain threshold value and have the same sign, c) to the clusters (G ', G °) assigned trajectory points (TP) in each case a section of the local Conversely (αβγ) from the three-dimensional point cloud ( 3DP ) is determined, resulting in a juxtaposition of these sections the video. Method according to claim 1, characterized in that in order to form the three-dimensional point cloud ( 3DP ) several scans with different locations of the laser scanner ( 10 ) are each generated with a center (C _i ), which are registered in the common coordinate system (XYZ) of the scene. A method according to claim 2, characterized in that the centers (C _i ) of the plurality of scans belong to the marked points. Method according to one of the preceding claims, characterized in that the sequence of trajectory points (TP) in the trajectory (T) on the basis of sorting criteria, in particular the spatial proximity of the excellent points and / or the temporal sequence of the scans, or by means of an optimization is defined. Method according to one of the preceding claims, characterized in that in a third step for each cluster (G ', G °) at least one focus (F) is defined, wherein for each "extended" cluster (G °) the focus (F) is preferably the center of gravity of the cluster (G °) and for each "even" cluster (G ') the focus (F) is preferably the end point of the cluster (G'). A method according to claim 5, characterized in that in a fourth step at each predetermined point of the trajectory (T), the three-dimensional point cloud ( 3DP ) with a preferred viewing direction on the focus (F) and the visible section of the local environment (αβγ) is determined. Method according to one of the preceding claims, characterized in that in the video, the trajectory points (TP) are approached in succession and in each case the section of the local environment (αβγ) by means of a display device ( 24 ) will be shown. Laser scanner ( 10 ) for carrying out a method according to claim 7, comprising a foot ( 14 ) and one relative to the foot ( 14 ) rotatable measuring head ( 12 ) with a light transmitter ( 17 ), which transmits a transmitted light beam ( 18 ), a light receiver ( 21 ), one of an object (O) in the vicinity of the laser scanner ( 10 ) reflected or otherwise scattered received light beam ( 20 ), and a control and evaluation device ( 22 ), which determines at least the distance to the object (O) for a plurality of measurement points (X) of the scan, registers the generated scans with respect to a common coordinate system (XYZ) of the scene and to each of the trajectory points (TP) the local environment (αβγ) from the three-dimensional point cloud ( 3DP ), wherein the display device ( 24 ) to the control and evaluation device ( 22 ) connected.

Images